Simulink 控制VR环境中的小车。小车有5个距离传感器，能够慢慢学会避开墙壁和障碍物。小车采用加强学习（Q learning），采用神经网络对Q函数逼近。由于使用了模拟退火，小车在开始的时候会经常撞击障碍物，10次后基本就不会再撞了。

小车的外观模型使用了"w198406141"在本论坛的虚拟现实区发布的VR模型。-VR environment Simulink control car. There are 5 car distance sensor, can gradually learn to avoid walls and obstacles. Car used to enhance learning (Q learning), using neural networks Q function approximation. As the use of simulated annealing, the car will start when the regular crash barrier, 10 times the fundamental will no longer hit. The appearance model used car " w198406141"　in this forum area of virtual reality VR model release.